EP0649225A1 - Method for controlling a bridge circuit and use of the method in the operation of an AM-broadcast transmitter - Google Patents

Abstract

A method for controlling the switches (S1-S4) of a bridge circuit (1) is indicated, in which method, if no voltage is intended to be applied to the load (ZL), the switches are short-circuited via the load. In every case, this provides a defined potential via the load, and the source impedance, as evidenced by the load, is essentially constant. Bridge circuits controlled with the method according to the invention can therefore also be used for applications which require a defined and constant source impedance. <IMAGE>

Description

Technical field

The invention relates to the field of power electronics. It relates to a method for controlling a bridge circuit according to the preamble of claim 1.

It also relates to an application of the method for operating an amplitude-modulated radio transmitter.

State of the art

A method of the type mentioned is described, for example, in the textbook "Power Electronics" by R. Lappe et al., Springer Verlag 1988.

The bridge circuit comprises two half bridges, each with two switches connected in series. The common nodes of the switches of each half-bridge form the load connections. The four switches are now controlled so that two diagonally opposite switches always conduct, while the others are open. As a result, either the positive or the negative pole of a DC voltage source, which can be connected to the bridge circuit, can be switched through to the load. As a result, the voltage on the load changes its sign.

Such bridge circuits and methods for controlling them are widely used as inverters.

From the article "Kilowatts on order" by R.S. Burden, IEEE Spectrum February 1993, further applications of these bridge circuits are now known. As pulse duration modulation amplifiers, they are used for all possible purposes such as high-performance function generators, broadband AC sources, etc. For this purpose, the switches of the bridge circuit are controlled with pulse duration modulated signals. In the pulse breaks e.g. of a pulse duration modulated output signal, all switches are opened and the load current e.g. an inductive load can flow off via free-wheeling diodes, which are parallel to the switches.

If all switches are open, however, the source impedance that the load sees is not defined. If no free-wheeling diodes are provided in the bridge circuit, not even the potential of the load is defined.

However, a source impedance that is defined at all times is required especially for applications in which a passive filter, for example a low pass, is connected downstream of the bridge circuit. However, a defined, as small as possible source impedance would also be advantageous for an inductive load to which a signal with a defined ripple is to be output. With a smaller source impedance, the time constant R / L (essentially R = internal resistance of the source and L = load inductance) would be smaller, and thus the steepness of any ripple would be lower. In this way, the switching frequency for a required, maximum ripple component could be significantly reduced.

In the prior art driving methods, however, the source impedance in the state where all the switches are open is much higher than in the other two states. A bridge circuit for applications that require an approximately constant source impedance is therefore ruled out from the outset.

Presentation of the invention

The object of the present invention is therefore to specify a method for controlling a bridge circuit in which the source impedance is defined at all times and is essentially constant, and to indicate an application of the method.

This object is achieved in a method of the type mentioned at the outset by the features of claim 1.

The essence of the invention is therefore that the switching state, in which all switches are open and there is no voltage at the output, is replaced by states in which the load is short-circuited.

In a preferred embodiment, freewheeling diodes are provided in parallel with the switches, and a short circuit across the load can also be achieved by only closing one switch depending on the current direction.

The advantage of the construction according to the invention is in particular that the potential of the load and the source impedance seen by the load are defined in every state. A bridge circuit controlled according to the invention can therefore also be used for applications which require a constant source impedance, for example for a downstream passive filter. It also becomes the ripple-determining one Time constant smaller, and a lower switching frequency is sufficient for an output voltage with a defined ripple component. It is also possible to connect the outputs of several bridge circuits in series and thus achieve a higher output voltage.

The application of the method according to the invention to the operation of an amplitude-modulated radio transmitter which comprises a plurality of similar RF stages of the same type which can be switched on and off, each of which is designed as a bridge circuit and has an AC signal of the same frequency corresponding to the carrier frequency at its outputs, and emit predominantly the same and constant amplitude, the output signals of the RF stages being added up and emitted via a filter to an antenna, and the number of switched-on stages being changed in accordance with an LF input signal, characterized in that at least some of them of the HF stages is controlled by the method according to the invention.

In a first preferred embodiment of this application, the control of the HF stages according to the method according to the invention is used to change the amplitude of the output signals of the relevant stages by the envelope of the carrier oscillation resulting from the summation in order to better approximate the LF input signal.

In a second preferred embodiment of this application, all RF stages are controlled according to the method according to the invention, and all RF stages are controlled in one of the states 31 and 32, respectively, if an interference voltage endangering the RF stages occurs before and / or after the filter.

Further embodiments result from the dependent claims.

Brief description of the drawings

The invention is explained in more detail below on the basis of exemplary embodiments in conjunction with the drawings.

Fig. 1

Ein Schaltbild einer Brückenschaltung;

Fig. 2

Ein Schaltbild einer Brückenschaltung mit Freilaufdioden;

Fig. 3

ein Blockschaltbild eines amplitudenmodulierten Rundfunksenders mit einzelnen HF-Stufen in Form von Brückenschaltungen, bei welchem Rundfunksender das erfindungsgemässe Verfahren Anwendung findet; und

Fig. 4

Zeitdiagramme für die Steuerung der Schalter S1,..,S4 einer Brückenschaltung gemäss Fig. 1 bei einem herkömmlichen Rundfunksender gemäss der US-A-4,580,111 (Fig. 4a) und gemäss einem bevorzugten Ausführungsbeispiel für die Anwendung des erfindungsgemässen Verfahrens bei einem AM-Rundfunksender (Fig. 4b).

Show it:

Fig. 1

A circuit diagram of a bridge circuit;

Fig. 2

A circuit diagram of a bridge circuit with free-wheeling diodes;

Fig. 3

a block diagram of an amplitude-modulated radio transmitter with individual RF stages in the form of bridge circuits, in which radio transmitter the inventive method is used; and

Fig. 4

1 for a conventional radio transmitter in accordance with US Pat. No. 4,580,111 (FIG. 4a) and in accordance with a preferred exemplary embodiment for the use of the method according to the invention in an AM radio transmitter (Fig. 4b).

The reference symbols used in the drawings and their meaning are summarized in the list of designations. In principle, the same parts are provided with the same reference symbols in the figures.

Ways of Carrying Out the Invention

Figure 1 shows the circuit diagram of a bridge circuit (1). It comprises a DC voltage source (2), which can be formed, for example, by a rectifier circuit with a supporting capacitor and outputs a voltage Uo, and at least two half bridges (3.1, 3.2), each parallel to the DC voltage are switched. The half bridges (3.1, 3.2) in turn comprise two switches connected in series (S1 and S3 or S2 and S4). The common node of the switches forms a load connection (4.1, 4.2). The switches are now controlled by a control unit (not shown) in such a way that, via a load (ZL) that can be connected to the connections (4.1 and 4.2), either the positive DC voltage Uo, the negative or no voltage occurs.

Switches S1 and S4 are closed for a positive load voltage, switches S2 and S3 for a negative one. If no voltage is to be applied to the load, all switches are left open according to the prior art. This results in the following 3 switching states: Status S1 S2 S3 S4 1 to open open to 2nd open to to open 3rd open open open open

In state 3 the voltage across the load is zero, but the potential of the load is not defined, the load "floats". Likewise, the source impedance seen by the load is theoretically infinite. However, as stated above, this is very undesirable.

According to the invention, state 3 can now be circumvented by replacing it with the following states: Status S1 S2 S3 S4 31 to to open open 32 open open to to

The load (ZL) is thus short-circuited in states 31 and 32 via switches S3 and S4 or S1 and S2, load potential and source impedance are defined, and the load voltage is zero as required.

An inductive load stores energy that would like to flow away via the short circuit. If the switches (S1-S4) conduct backwards, this can be done without any special measures. However, if they can only carry current in one direction, freewheeling diodes (D1-D4) must be provided in parallel with the switches ( Figure 2 ). This allows the load current to flow through a free-wheeling diode and the closed switches.

If the load current flows from the first load connection (4.1) to the second (4.2), the switch S2 in state 31 is not necessary to carry the current, so that it can also be left open. The same applies to switch S3 in state 32. If, on the other hand, the current flows in the opposite direction, switch S1 can be left open in state 31 and switch S4 in state 32.

When switching with free-wheeling diodes, the source impedance is of course also defined in state 3. With the control according to the invention, i.e. the states 31 and 32, however, this is significantly reduced and is therefore approximately constant over all states.

With the control method according to the invention, the potential of a load of a bridge circuit is therefore defined in every switching state. A bridge circuit which is controlled according to the invention can therefore also be used for applications which require a defined load potential and thus a defined source impedance (for example downstream filters). In addition, the source impedance is kept fairly constant at a small level. This has the great advantage that a lower switching frequency can be used for a required ripple height, because the dominant time constant R / L (essentially: R = source impedance, L = load inductance) is much smaller. As a result, the ripple rises less steeply and is smaller overall. Another advantage is that several bridge circuits can be connected in series.

According to the invention, the control method described above can be used with particular advantage when operating an amplitude-modulated radio transmitter 14 constructed with bridge circuits, as is shown in the block diagram in FIG. 3. Such a transmitter, the basic principle of which is e.g. in the publications US-A-4,580,111 (or EP-B1-0 083 727) or US-A-4,859,967, comprises a plurality of RF stages 1.1, .., 1.N, each internally as a bridge circuit of the are constructed in Fig. 2 shown. Power MOSFETs are preferably used as switches S1, .., S4.

The RF stages or bridge circuits 1.1, .., 1.N are controlled by a controller 6 from a carrier frequency source 5, the switches S1, .., S4 of the individual stages being switched so that a square-wave signal is present at the output of the stage results, whose fundamental frequency corresponds to the carrier frequency. The outputs of the HF stages 1.1,... 1.N are connected to a summation device 7, which has a corresponding number of primary windings 12 and one or more secondary windings 13. The primary winding 12 in each case corresponds to the load impedance ZL from FIG. 1 or FIG. 2.

The output signals from the RF stages 1.1,..., 1N, which are summed up in the summing device 7 and which are initially approximately square-wave signals, are freed from the harmonics in a subsequent filter 9, which is designed, for example, as a bandpass filter, and onto an antenna 11 given and emitted. The amplitude modulation is achieved in that according to the LF input signal from the controller 6 present at an LF input 16, only as many HF stages are used can be controlled as correspond to the instantaneous value of the AF amplitude. Details of the control can be found in the above-mentioned documents.

Various problems arise with such a (digitally operating) AM radio transmitter: firstly, when using similar RF stages, the constantly varying LF signal in the amplifier / modulator is approximated by a staircase-shaped signal. The quality of the approximation depends on the number of stages, so that a high circuit outlay would have to be carried out for a good approximation. On the other hand, interference voltages, which are caused at the transmitter output due to mismatching to the antenna or lightning strike, can act back on the individual RF stages 1.1,..., 1.N and their MOSFET switches via the filter 9 and the summation device 7 and lead to destruction .

As a solution to the first problem, US-A-4,580,111, which has already been mentioned, has proposed a solution to use, in addition to the similar RF stages, further stages which emit only a fraction of the output amplitude of the other stages and are preferably graduated in binary form. However, this has the disadvantage that different types of RF stages must be used in the transmitter with regard to the power range, which leads to increased system costs. It is therefore more favorable to design all RF stages identically and to achieve intermediate values in the approximation by means of a special control of individual bridge circuits or of their switches S1, .., S4.

This principle can be explained in more detail with reference to FIG. 4: FIG. 4a shows the time diagrams of the known type of bridge control in an AM radio transmitter of the type described. The two half bridges (3.1 and 3.2 in FIG. 1) with their switches S1, S3 and S2, S4 are switched synchronously, with switches S1 and S4 switched on (ON) and switches S2 and S3 switched off (OFF) are, and vice versa. There is always a DC voltage at the bridge output, which changes its sign in the rhythm of the switching processes, so that a rectangular AC output voltage U1 results. The RF stages either all have the same phase and the same amplitude or - if the fine approximation according to US-A-4,580,111 is used - partially a binary graduated amplitude.

With the RF level unchanged, intermediate values can now be achieved with much less effort by having at least one level with the same maximum amplitude, i.e. at the same DC voltage Uo, the positive and negative half-waves emitted by the bridge circuit at the output are changed in their pulse duration. The shorter the pulse duration, the smaller is the amplitude of the amplitude of the fundamental oscillation present at the transmitter output due to the temporal integration effect of the filter, so that the intermediate values between two approximation stages can be set continuously by a constant change in the pulse duration.

According to FIG. 4b, the pulse duration can be set simply in that the half bridges 3.1 and 3.2 are no longer controlled synchronously, but rather out of phase with one another, ie the switches S1 and S3 of the first half bridge 3.1 are activated at times t1, t3, t5, .. ., which are different from the switching times t2, t4, ... of the switches S2 and S4 of the second half bridge 3.2. An alternating output voltage U2 then results, which consists of positive and negative square-wave pulses, which have the same amplitude as the voltage U1 in FIG. 4a, but have a pulse duration that can be varied in accordance with the phase shift between the half-bridges.

In this context, one immediately recognizes that with this type of switch control, in the time intervals (t1-t2, t3-t4, ...) in which the voltage at the bridge output (U2) is zero, either the upper switches S1, S2 or the lower Switches S3, S4 are switched on, so that the method according to the invention is used in this type of bridge control. The associated short circuit of the primary windings 12 of the summing device 7 ensures that the filter 9 sees a finite source impedance in any case, without which a proper function of the filter is not possible.

The second problem is in the above No. 4,859,967 has been proposed to detect the interference voltages at the output and to switch off the HF stages when such an interference voltage occurs and, at the same time, to apply a reverse voltage, which is in phase with the interference voltage, to the MOSFETs in the bridge circuits via the driver stage in order to destroy destructive potential differences at the MOSFETs to prevent. This requires additional electronics, which is complex and must be designed very precisely with regard to the phase relationships.

In contrast, the influence of interference voltages at the transmitter output on the RF stages 1.1, .., 1.N can be limited in a significantly simpler and less critical manner if, according to a further application of the method according to the invention, all RF stages 1.1, .., 1.N in one of the states 31 and 32 are controlled if an interference voltage endangering the RF stages 1.1, .., 1.N occurs before and / or after the filter 9. As a result, the primary windings 12 of the summation device are short-circuited, so that no voltages dangerous for the bridge circuits 1 can be induced in them.

In order to detect an interference voltage endangering the RF stages 1.1, .., 1.N, the phase between the current and the voltage and / or behind the filter 9 is preferably measured in front of the filter 9 by means of a phase measuring device 8 (FIG. 3) by means of a reflection measuring device 10 measured the antenna reflection factor. If a fault occurs, a corresponding signal is sent to the controller 6 via a control line 15, which then immediately the RF stages 1.1, .., 1.N in the Short circuit controls, ie switches S1 and S2 and / or S3 and S4 in each bridge circuit 1 closes.

Label list

1

Bridge circuit

1.1, .., 1.N

RF level

2nd

DC voltage source

3.1, 3.2

Half bridges

4.1, 4.2

Load connections

5

Carrier frequency source

6

control

7

Summation device

8th

Phase measuring device

9

Filter (band pass)

10th

Reflection measuring device

11

antenna

12th

Primary winding

13

Secondary winding

14

Radio station

15

Control line

16

NF input

S1-S4

switch

ZL

Load impedance

D1-D4

Free wheeling diodes

Uo

DC voltage

U1, U2

Output voltage

t1, .., t5

time

Claims (6)

Method for controlling a bridge circuit (1) which comprises at least two half bridges (3.1, 3.2), each with two switches S1 and S3 or S2 and S4, the common nodes of the switches (S1, S3 or S2, S4) each Form a load connection (4.1, 4.2) to which a load (ZL) can be connected, using which method the switches (S1-S4) are controlled repeatedly according to one of the following two switching states: Status S1 S2 S3 S4 1 to open open to 2nd open to to open characterized in that the switches (S1-S4) are controlled in a further state in which no voltage is supplied to the load (ZL) in such a way that one of the switching states in the following table results: Status S1 S2 S3 S4 31 to to open open 32 open open to to A method according to claim 1, characterized in that a) free-wheeling diodes (D1-D4) are provided in parallel with the switches (S1-S4), b) the switch (S2) in state 31 and the switch (S3) in state 32 can be left open, if a current flows through the load (ZL), from the load connection (4.1) to the load connection (4.2), c) the switch (S1) in state 31 and the switch (S4) in state 32 can be left open if a current flows through the load (ZL) from the load connection (4.2) to the load connection (4.1). Application of the method according to claim 1 to the operation of an amplitude-modulated radio transmitter (14) which comprises a plurality of similar RF stages (1.1, .., 1.N) which can be switched on and off individually and which are each designed as a bridge circuit (1) are and emit at their outputs an alternating voltage signal of the same frequency, the carrier frequency corresponding, and predominantly the same and constant amplitude, the output signals of the RF stages (1.1, .., 1.N) added up and a filter (9) to one Antenna (11) are emitted, and the number of switched-on stages is changed in accordance with an LF input signal, characterized in that at least some of the HF stages (1.1, .., 1.N) are controlled according to the method of claim 1 becomes. Use according to claim 3, characterized in that for better approximation of the LF input signal by the envelope of the carrier oscillation resulting from the summation, the half bridges (3.1, 3.2) of the at least part of the HF stages (1.1, .., 1.N ) are driven out of phase with each other, and that in the time intervals (t1-t2, t3-t4, ...) in which the output voltage (U2) of the at least some of the HF stages (1.1, .., 1.N) is zero, either the upper switches S1, S2 or the lower switches S3, S4 are switched on. Use according to claim 3, characterized in that all RF stages (1.1, .., 1.N) are controlled by the method according to claim 1, and that all RF stages (1.1, .., 1.N) can be controlled in one of the states 31 or 32 if an interference voltage endangering the HF stages (1.1, .., 1.N) occurs before and / or after the filter (9) . Use according to claim 5, characterized in that in order to detect an interference voltage endangering the HF stages (1.1, .., 1.N) in front of the filter (9), the phase between the current and the voltage and / or behind the filter (9 ) the antenna reflection factor can be measured.

Images